"Sam" <[EMAIL PROTECTED]> wrote:
> > This part is not true. You only need to receive the _header_, and can
> > still return the appropriate rejection code after your check. This
>
> Nope. You cannot receive just the header, send an SMTP rejection,
> then expect the sender to stop sending you the rest of the
> message...
Conceded. You win your point in favor of per-host blocking. However, I
still think my suggestion (made satirically) beats any other
pattern-matching solution which involves scanning headers or otherwise
"sucking in the entire message".
If it were widely adopted, the spam would be sent out in 24-recipient
blocks...then 15, then 9, then 5, then 1 as mail admins reacted to the
countermeasure. Ultimately, the whole Internet might become a BCC-free
zone, yet spam would still go through.
Does any non-spammer routinely include >25 (or even >5) BCC's in a
message? The only exception I can think of is corporate email, which,
of course, is immune to such rules since the corporate mail server can
handle them appropriately.
> > Sam Varshavchik has a patch to qmail-smtpd which calls procmail
> > recipes to filter spam before it is accepted for delivery.
>
> A slightly outdated description, but that's not important.
Oops--I told you about your own patch. That's pretty funny!
Len.
--
The king by judgment establisheth the land: but he that receiveth
gifts overthroweth it. --Proverbs 29:4
